Background: Large observational studies have demonstrated a clear inverse association between renal function and risk of aortic stenosis (AS). Whether this represents a causal, reverse causal or correlative relationship remains unclear. We investigated this using a bidirectional 2-sample Mendelian randomization approach.
Methods and results: We collected summary statistics for the primary analysis of chronic kidney disease (CKD) and AS from genome-wide association study meta-analyses including 480 698 and 653 867 participants, respectively. We collected further genome-wide association study summary statistics from up to 1 004 040 participants for sensitivity analyses involving estimated glomerular filtration rate (eGFR) derived from creatinine, eGFR derived from cystatin C, and serum urea nitrogen. Inverse-variance weighted was the primary analysis method, with weighted-median, weighted-mode, Mendelian randomization-Egger, and Mendelian randomization-Pleiotropy Residual Sum and Outlier as sensitivity analyses. We did not find evidence of a causal relationship between genetically predicted CKD liability as the exposure and AS as the outcome (odds ratio [OR], 0.94 per unit increase in log odds of genetic liability to CKD [95% CI, 0.85-1.04], P=0.26) nor robust evidence of AS liability as the exposure and CKD as the outcome (OR, 1.04 per unit increase in log odds of genetic liability to AS [95% CI, 0.97-1.12], P=0.30). The sensitivity analyses were neutral overall, as were the analyses using eGFR derived from creatinine, eGFR derived from cystatin C, and serum urea nitrogen. All positive controls demonstrated strong significant associations.
Conclusions: The present study did not find evidence of a substantial effect of genetically predicted renal impairment on risk of AS. This has important implications for research efforts that attempt to identify prevention and treatment targets for both CKD and AS.
Keywords: Mendelian randomization; aortic stenosis; genetics; renal function; valve disease.